HCLSIG/SWANSIOC/MemberSubmission

From W3C Wiki

Things to emphasise in the document:

  • Bridging Web 2.0 and BioInformatics efforts
  • OWL-DL version of SIOC
  • SCF framework (needs more work b4 submitting)

SWANSIOC Member Submission

Abstract

SWANSIOC is a shared effort motivated by the use social software by researchers in biomedicine. It aims to provide a common framework for scientific discourse representation using Semantic Web technologies.

Status

@@todo

Introduction

This document introduces the SWAN/SIOC ontological alignment work, as well as a related applications designed to help researchers to better share ideas around particular scientific topics.

Motivations for the SWANSIOC project

Online community sites in general (forums, weblogs, bulletin boards, etc.) have replaced the traditional means of keeping a community informed via libraries and publishing. They are a valuable source of information and quite often it is a community site where you would end up when searching for some information. But there is a problem - online community sites are like islands without bridges connecting them. You may find information in a forum, but not know that there are missing pieces of related information that can be found on other community sites.

SIOC (Semantically-Interlinked Online Communities) is an attempt to link online community sites and to use Semantic Web technologies to describe the information community sites have about their structure and contents. An aim of SIOC is to allow people to find related information in other online communities and to discover new connections between discussion posts. The SIOC project is a sub-initiative of the DERI Líon project (funded by SFI).

In parallel with the SIOC effort, researchers are now beginning to realise the potential of social web technologies for scientific, legislative and other domain-specific discourses. Both formal scientific works in publications and also research discourse in community mechanisms can and should be interlinked with semantics. For example, efforts like bio-zen and SISC are aim ing to represent data, information and knowledge from research in all facets of the life sciences on the Semantic Web.

There is a need to provide structured representations of professional scientific discourse for the HCLS domain, and this fits well with a future direction of SIOC to augment the existing framework with terms and applications specific to various domains.

Alzheimer Disease (AD) and Parkinson Disease (PD) are devastating neurodegenerative disorders for which there is no cure, and whose mechanisms (etiology) are incompletely understood. There are currently some 5 million Alzheimer patients and more than 1 million Parkinson patients in the U.S. alone, with the cost of care-giving running into the hundreds of billions of dollars. These numbers are expected to double over the next several decades because of projected increases in the aged population. AD is now the third most expensive disease to treat in the U.S., costing society close to $100 billion annually.

AD and PD are characterized by the loss of function and eventual death of massive numbers of neurons, beginning in specific brain regions (entorhinal cortex in AD and substantia nigra in PD).

There are a number of ways in which closer integration of stem cell, AD and PD research could be beneficial:

  • Emerging hypotheses propose that certain molecules that play a central role in AD and PD are involved in neural regeneration, and that a possible cause of cell death is the loss of this regenerative capacity. These molecules include nerve growth factor, amyloid precursor protein and dopamine.
  • Maintaining stable cultures of human neurons has been difficult, and has impeded progress in developing "test tube" models to test hypotheses and screen drugs. Using embryonic stem cells to generate human neurons and other types of brain cells could lead to better test-tube models of neurodegenerative disease.
  • Therapy development is exceptionally challenging for neurodegenerative diseases because in the adult central nervous system, neurons generally are not capable of regenerating to replace diseased and dying cells. Stem cells may be manipulated to develop cell lines and engineered tissue suitable for transplantation therapy.
  • Stem cell biology may provide knowledge to harness the brain's innate regenerative capacity for therapeutic purposes.

Alzheimer Disease research is the focal area of the oldest and largest biomedical web community by and for AD researchers, Alzforum (www.alzforum.org).

The SWAN ontology and knowledgebase (Ciccarese et al. 2008, Journal of Biomedical Informatics, in press) is a joint project of the Alzforum and the Massachusetts General Hospital. Stem Cell technology is likewise the subject of StemBook, an online publication of the Harvard Stem Cell Institute. StemBook is implemented using the Science Collaboration Framework (SCF), a special distribution of Drupal which among other capabilities can node-proxy resources on SPARQL endpoints (lazily instantiated node data) and understands certain elements of SWAN. A third web community, PD Online Research, also based on SCF, is now under development with scheduled deployment in Spring 2009 and planned integration with SWAN.

These communities with intersecting but distinct research interests are poster children for semantic interoperability of discourse. They form a convenient and perhaps ideal driving biological project for integrating SWAN and SIOC while keeping requirements grounded in the needs of actual biomedical researchers.

Namespaces

Prefixes and associated namespaces used in this document:

@@TODO

The SWANSIOC Ontology

Introduction

SIOC OWL-DL Compliance

  • SIOC was RDFS + bits of OWL (disjointWith) => owl:full
  • SWAN uses OWL inferencing
  • Need to adapt SIOC to OWL-DL / Lite
  • Improvements to make it OWL-DL
  • Future work: focus on constraints (e.g. cardinality), etc to validate the model.

SWAN v1.2

Alignment with SWAN and SIOC

  • Keep SIOC lightweight
  • Add most of the alignemnets in a SWANSIOC module
  • Introduce re-usable terms in the SIOC Type module
  • list the aligned properties / classes here

Applications

Discussions

future works regarding scientific discourse effort in HCLS

Ack

References